langchain-ai/deepagents · error · ValueError

Hook resume invocation_id mismatch: expected {invocation_id}

Error message

Hook resume invocation_id mismatch: expected {invocation_id}, got {response.invocation_id}

What it means

parse_hook_resume_value validates that a hook resume payload (submitted when the user resolves a hook interrupt) carries the same invocation_id as the suspended invocation. The pydantic HOOK_INVOCATION_RESPONSE_ADAPTER first parses the raw value; then this check ensures the response corresponds to the invocation being resumed. A mismatch means the resume value belongs to a different hook invocation and is rejected with ValueError.

Source

Thrown at libs/code/deepagents_code/hooks/interrupt.py:107

    Args:
        value: Resume payload returned by the client.
        invocation_id: Expected invocation id from the request.
        snapshot_id: Expected configuration snapshot id.

    Returns:
        Validated response.

    Raises:
        ValueError: If the resume payload is missing, mistyped, or mismatched.
    """
    response = HOOK_INVOCATION_RESPONSE_ADAPTER.validate_python(value)
    if response.invocation_id != invocation_id:
        msg = (
            f"Hook resume invocation_id mismatch: expected {invocation_id}, "
            f"got {response.invocation_id}"
        )
        raise ValueError(msg)
    if response.snapshot_id != snapshot_id:
        msg = (
            f"Hook resume snapshot_id mismatch: expected {snapshot_id}, "
            f"got {response.snapshot_id}"
        )
        raise ValueError(msg)
    return response


def is_hook_interrupt_payload(value: object) -> bool:
    """Return whether `value` looks like a Hooks v2 invocation interrupt."""
    return (
        isinstance(value, dict) and value.get("type") == HOOK_INVOCATION_INTERRUPT_TYPE
    )

View on GitHub (pinned to a1af029e6e)

Solutions

1. Re-read the pending interrupt payload and resume with its exact invocation_id (and snapshot_id)
2. Discard stale resume values and re-request the interrupt — the old invocation_id is no longer valid
3. If multiple interrupts are pending, key resume payloads by interrupt id so the right value reaches the right invocation
4. Check for framework updates if interrupt IDs are regenerated on retry, and upgrade to a version that preserves them

Example fix

// before
resume_value = saved_resume  # from an earlier session
result = parse_hook_resume_value(saved_resume, expected_invocation_id, snapshot_id)

// after
pending = engine.get_pending_interrupts()[expected_invocation_id]
resume_value = pending.build_resume(user_input)
result = parse_hook_resume_value(resume_value, expected_invocation_id, snapshot_id)

from deepagents_code.hooks.interrupt import parse_hook_resume_value

# verify pairing before resuming
if resume_payload.get("invocation_id") != expected_invocation_id:
    raise ValueError("stale resume payload: refetch pending interrupt")

def matches_invocation(payload: dict, invocation_id: str) -> bool:
    return payload.get("invocation_id") == invocation_id

try:
    resume = parse_hook_resume_value(payload, invocation_id, snapshot_id)
except ValueError as exc:
    logger.error("hook resume rejected: %s", exc)
    resume = refetch_and_resume(invocation_id)  # re-answer the fresh interrupt

When it happens

Trigger: Resuming a hook interrupt via _invoke_hook/parse_hook_resume_value with a payload whose invocation_id differs from the expected one — e.g. resuming an old interrupt after the agent re-issued a new hook invocation, or mixing up interrupt IDs when several hooks are pending.

Common situations: Stale UI state in the coding agent (TUI) submitting an outdated resume after re-render; manually replaying saved resume payloads across sessions; concurrent hook interrupts where the caller pairs the wrong payload with the wrong invocation.

Related errors

• Hook resume snapshot_id mismatch: expected {snapshot_id}, go
• Permission interrupted by hook
• Received hook invocation interrupt without a HooksRuntime
• Failed to parse hook interrupt
• -32600

AI-assisted analysis of langchain-ai/deepagents@a1af029e6e (2026-08-29). Data as JSON: /api/errors/e31a26f6219ff116. Report an issue: GitHub.